The Impact of Two-Invoice System on Pharmaceutical Manufacturers’ Selling Expenses in China: A Difference-in-Differences Approach

Ran Yi, Yuanyuan Hu, Shouming Chen, Fangjun Qiu, Ahmed RabeeuInternational Journal of Environmental Research and Public Health 2022 · DOI

TL;DR

China's "Two-Invoice System" (TIS) for drug procurement, implemented in 2017, aimed to curb excessive drug prices. This study finds that, counterintuitively, the TIS initially increased pharmaceutical manufacturers' selling expenses, particularly for those using an agency-based distribution model. While this effect diminishes over time, it highlights a short-term cost shift that warrants further policy consideration.

What Problem Does This Paper Address?

The pharmaceutical industry globally grapples with the issue of high drug prices, which impacts affordability and access to healthcare. In China, a significant portion of drug costs is attributed to complex distribution chains and promotional activities. To address this, China introduced the Two-Invoice System (TIS) in 2017, a reform designed to streamline drug circulation by limiting the number of invoices to two: one from the manufacturer to the distributor, and another from the distributor to the medical institution. The intended outcome was to increase price transparency, reduce the number of intermediaries, and ultimately lower drug prices. However, the authors question whether this reform might have unintended consequences, specifically on the selling expenses of pharmaceutical manufacturers. They investigate if the TIS, by compressing distribution channels, forces manufacturers to internalize more marketing and sales functions, thereby increasing their own selling costs.

Methods

The core of this study employs a Difference-in-Differences (DID) methodology. This approach is well-suited for evaluating the impact of a policy change (like the TIS) by comparing the outcomes of a "treatment group" (firms affected by the policy) to a "control group" (firms not affected or less affected) before and after the policy's implementation.

Here's a breakdown:

  • Treatment Group: Pharmaceutical manufacturers directly impacted by the TIS.
  • Control Group: Pharmaceutical manufacturers whose primary business revenue comes from areas less affected by the TIS, such as medical machinery production or other devices.
  • Time Periods: Data from 2014-2020 is used, with 2017 being the year of initial implementation and 2018 onwards considered the post-treatment period. The year 2017 was excluded from the main analysis to avoid the transitional effects of policy adoption.
  • Dependent Variable: The natural logarithm of selling expenses (InSE), which includes costs related to distribution, marketing, and sales.
  • Key Independent Variable: The DID interaction term, which captures the combined effect of being in the treatment group and being in the post-treatment period.
  • Control Variables: A range of firm-level characteristics (e.g., ROA, Tobin's Q, leverage, size, age) are included to account for other factors that might influence selling expenses.
  • Robustness Checks: The authors conduct several robustness checks, including:
  • Dynamic DID: Examining the effect over multiple periods before and after the TIS implementation to assess the temporal dynamics.
  • Placebo Test: Advancing the policy implementation date by three years to see if the effect disappears, confirming the TIS is the cause.
  • Alternative Dependent Variable: Using selling expenses intensity (selling expenses/sales) to account for potential firm-size effects.
  • Propensity Score Matching (PSM-DID): Matching treated firms with similar control firms to mitigate unobservable differences.
  • Excluding 2020: Removing data from the COVID-19 year to check for confounding effects.

Datasets

The study utilizes data from A-share listed pharmaceutical manufacturing firms on the Shanghai Stock Exchange and Shenzhen Stock Exchange. The data spans from 2014 to 2020.

  • Accounting statement and selling expenses data: Sourced from the CSMAR database and Wind databases.
  • TIS implementation data: Manually gathered from official government websites and other relevant sources to determine the regional adoption timeline of the TIS.

No new benchmarks were introduced; the study relies on existing financial databases.

Key Insights and Ideas

  • The TIS is a double-edged sword for manufacturers' selling expenses: While intended to lower drug prices, the immediate effect of the TIS is an increase in selling expenses for pharmaceutical manufacturers. This suggests that the cost savings from reduced intermediaries are not fully realized by manufacturers in the short term.
  • Distribution model matters significantly: The impact of the TIS is not uniform. Manufacturers relying on an agency mode (AM), where distributors handle much of the marketing and promotion, see a substantial increase in their selling expenses. In contrast, those using a direct sales office model (DSOM) experience little to no significant change.
  • Short-term pain for long-term gain? The study suggests that the increase in selling expenses is temporary, diminishing over time. This implies that the TIS might eventually lead to lower drug prices and a more efficient market, but at an initial cost to manufacturers.
  • Hidden costs of channel compression: The TIS forces manufacturers to take on more direct responsibility for marketing and sales functions, which were previously outsourced to a complex network of distributors. This "hidden selling expense" is now being brought to the forefront.

Key Results

  • The DID analysis shows a statistically significant positive impact of the TIS on the natural log of selling expenses. Specifically, the coefficient for the DID term indicates an increase in selling expenses following the TIS implementation.
  • In the agency mode (AM) group, the TIS led to a significant increase in selling expenses. For example, in one model, the DID coefficient was 0.3506 (p < 0.1).
  • In the direct sales office model (DSOM) group, the DID coefficients were insignificant or even negative, suggesting no significant increase in selling expenses due to the TIS.
  • The dynamic analysis indicates that the effect of the TIS on selling expenses is temporary, lasting approximately 12 to 24 months, and gradually diminishes over time.

Limitations and Future Work

  • Focus on direct impact: The study primarily examines the direct impact of the TIS on selling expenses. The authors acknowledge that the TIS likely has indirect effects on innovation investment by pharmaceutical manufacturers, which future research should explore.
  • Limited moderator variables: The analysis only considers the drug circulation mode as a moderator. Other factors, such as the mode of corporate value creation (marketing-orientation vs. innovation-orientation), could also influence the relationship and warrant further investigation.
  • Broader market impact: This study focuses solely on pharmaceutical manufacturers. The authors suggest that future research should consider the impact of the TIS on agents and intermediaries, such as Contract Sales Organizations (CSOs).
  • Tax-related risks: The study briefly mentions potential tax-related risks arising from the increased role of CSOs and the need for government measures like improving drug approval systems and promoting generic drug consistency evaluations. This could be an area for further policy-oriented research.

Summary updated

Abstract

A perennial question for the pharmaceutical industry has been excessive drug prices. To alleviate patients' burden of expensive medical bills and increase the affordability of medicines, China adopted the Two-Invoice System (TIS) in drug procurement for public medical institutions in 2017. In this paper, we study the impact of the TIS on pharmaceutical manufacturers' selling expenses. Using a Difference-in-Differences (DID) methodology and a sample of the A-share pharmaceutical manufacturing firms listed on the Shanghai Stock Exchange and Shenzhen Stock Exchange from the years 2014 to 2020, we find that the TIS leads to a significant increase in pharmaceutical manufacturers' selling expenses but gradually weakens over time. In addition, we further explore whether the impact of the TIS on pharmaceutical manufacturers' selling expenses is affected by the pharmaceutical manufacturers' previous drug circulation mode. The results indicate that the TIS could significantly increase the pharmaceutical manufacturers' selling expenses in the agency mode group. However, there is no evidence to support the TIS having the same effect in the direct sales office model group.
BibTeX
@article{yi2022impact,
  author = {Ran Yi and Yuanyuan Hu and Shouming Chen and Fangjun Qiu and Ahmed Rabeeu},
  title = {{The Impact of Two-Invoice System on Pharmaceutical Manufacturers’ Selling Expenses in China: A Difference-in-Differences Approach}},
  journal = {International Journal of Environmental Research and Public Health},
  year = {2022},
  doi = {10.3390/ijerph19074400},
}
Figure